Multi-decadal land cover change analysis over Sundarbans Mangrove Forest of Bangladesh: A GIS and remote sensing based approach

Abstract

Both natural processes and anthropogenic actions alter land use and land cover but the impact of anthropogenic actions has dramatically increased since the last century in many regions of the world. This study attempts to investigate the land cover change of the Sundarbans and surrounding areas of Bangladesh in the last forty years by analyzing Landsat images. Sundarbans, having the richest and exceptional biodiversity, is the largest remaining tract of mangrove forest on earth facing multidimensional threats of degradation. Time series Landsat satellite images collected for five years are 1980 (MSS), 1990 (TM), 2000 (TM), 2010 (TM) and 2020 (OLI). By analyzing, comparing and visualizing the classified images using ArcMap 10.5 software the historical trend of land cover change of the study area is exhibited. The result revealed that a large amount of non-vegetated areas located outside of Sundarbans mangrove forest converted into water bodies due to shrimp cultivation. During this long time, the Sundarbans mangrove forest area remained unchanged except for little change occurred by soil erosion along the river bank and coastline. Kappa coefficient revealed the accuracy of the land cover map with a coefficient value of 0.88, 0.92, 0.90, 0.93 and 0.92 for 1980, 1990, 2000, 2010 and 2020 images respectively. This research will support policymakers and the forest department to improve management policy. Also, influence researchers to conduct case-specific micro-scale research in the region. • Land cover change of Sundarbans Mangrove forest and surrounding area was analysed using GIS and RS techniques. • Excessive shrimp cultivation altered land cover outside of Sundarbans. • Sundarbans Mangrove forest remained unchanged over the last 40 years except for little change occurred by soil erosion. • Cyclone, sea level rise and decreased river water flow affecting the coastal island and the river bank. • Kappa coefficient value of 0.88, 0.92, 0.90, 0.93 and 0.92 found for 1980, 1990, 2000, 2010 and 2020 images respectively.